1============================================= 2Building a JIT: Per-function Lazy Compilation 3============================================= 4 5.. contents:: 6 :local: 7 8**This tutorial is under active development. It is incomplete and details may 9change frequently.** Nonetheless we invite you to try it out as it stands, and 10we welcome any feedback. 11 12Chapter 3 Introduction 13====================== 14 15Welcome to Chapter 3 of the "Building an ORC-based JIT in LLVM" tutorial. This 16chapter discusses lazy JITing and shows you how to enable it by adding an ORC 17CompileOnDemand layer the JIT from `Chapter 2 <BuildingAJIT2.html>`_. 18 19Lazy Compilation 20================ 21 22When we add a module to the KaleidoscopeJIT class from Chapter 2 it is 23immediately optimized, compiled and linked for us by the IRTransformLayer, 24IRCompileLayer and ObjectLinkingLayer respectively. This scheme, where all the 25work to make a Module executable is done up front, is simple to understand and 26its performance characteristics are easy to reason about. However, it will lead 27to very high startup times if the amount of code to be compiled is large, and 28may also do a lot of unnecessary compilation if only a few compiled functions 29are ever called at runtime. A truly "just-in-time" compiler should allow us to 30defer the compilation of any given function until the moment that function is 31first called, improving launch times and eliminating redundant work. In fact, 32the ORC APIs provide us with a layer to lazily compile LLVM IR: 33*CompileOnDemandLayer*. 34 35The CompileOnDemandLayer class conforms to the layer interface described in 36Chapter 2, but its addModuleSet method behaves quite differently from the layers 37we have seen so far: rather than doing any work up front, it just scans the 38Modules being added and arranges for each function in them to be compiled the 39first time it is called. To do this, the CompileOnDemandLayer creates two small 40utilities for each function that it scans: a *stub* and a *compile 41callback*. The stub is a pair of a function pointer (which will be pointed at 42the function's implementation once the function has been compiled) and an 43indirect jump through the pointer. By fixing the address of the indirect jump 44for the lifetime of the program we can give the function a permanent "effective 45address", one that can be safely used for indirection and function pointer 46comparison even if the function's implementation is never compiled, or if it is 47compiled more than once (due to, for example, recompiling the function at a 48higher optimization level) and changes address. The second utility, the compile 49callback, represents a re-entry point from the program into the compiler that 50will trigger compilation and then execution of a function. By initializing the 51function's stub to point at the function's compile callback, we enable lazy 52compilation: The first attempted call to the function will follow the function 53pointer and trigger the compile callback instead. The compile callback will 54compile the function, update the function pointer for the stub, then execute 55the function. On all subsequent calls to the function, the function pointer 56will point at the already-compiled function, so there is no further overhead 57from the compiler. We will look at this process in more detail in the next 58chapter of this tutorial, but for now we'll trust the CompileOnDemandLayer to 59set all the stubs and callbacks up for us. All we need to do is to add the 60CompileOnDemandLayer to the top of our stack and we'll get the benefits of 61lazy compilation. We just need a few changes to the source: 62 63.. code-block:: c++ 64 65 ... 66 #include "llvm/ExecutionEngine/SectionMemoryManager.h" 67 #include "llvm/ExecutionEngine/Orc/CompileOnDemandLayer.h" 68 #include "llvm/ExecutionEngine/Orc/CompileUtils.h" 69 ... 70 71 ... 72 class KaleidoscopeJIT { 73 private: 74 std::unique_ptr<TargetMachine> TM; 75 const DataLayout DL; 76 std::unique_ptr<JITCompileCallbackManager> CompileCallbackManager; 77 ObjectLinkingLayer<> ObjectLayer; 78 IRCompileLayer<decltype(ObjectLayer)> CompileLayer; 79 80 typedef std::function<std::unique_ptr<Module>(std::unique_ptr<Module>)> 81 OptimizeFunction; 82 83 IRTransformLayer<decltype(CompileLayer), OptimizeFunction> OptimizeLayer; 84 CompileOnDemandLayer<decltype(OptimizeLayer)> CODLayer; 85 86 public: 87 typedef decltype(CODLayer)::ModuleSetHandleT ModuleHandle; 88 89First we need to include the CompileOnDemandLayer.h header, then add two new 90members: a std::unique_ptr<CompileCallbackManager> and a CompileOnDemandLayer, 91to our class. The CompileCallbackManager member is used by the CompileOnDemandLayer 92to create the compile callback needed for each function. 93 94.. code-block:: c++ 95 96 KaleidoscopeJIT() 97 : TM(EngineBuilder().selectTarget()), DL(TM->createDataLayout()), 98 CompileLayer(ObjectLayer, SimpleCompiler(*TM)), 99 OptimizeLayer(CompileLayer, 100 [this](std::unique_ptr<Module> M) { 101 return optimizeModule(std::move(M)); 102 }), 103 CompileCallbackManager( 104 orc::createLocalCompileCallbackManager(TM->getTargetTriple(), 0)), 105 CODLayer(OptimizeLayer, 106 [this](Function &F) { return std::set<Function*>({&F}); }, 107 *CompileCallbackManager, 108 orc::createLocalIndirectStubsManagerBuilder( 109 TM->getTargetTriple())) { 110 llvm::sys::DynamicLibrary::LoadLibraryPermanently(nullptr); 111 } 112 113Next we have to update our constructor to initialize the new members. To create 114an appropriate compile callback manager we use the 115createLocalCompileCallbackManager function, which takes a TargetMachine and a 116JITTargetAddress to call if it receives a request to compile an unknown 117function. In our simple JIT this situation is unlikely to come up, so we'll 118cheat and just pass '0' here. In a production quality JIT you could give the 119address of a function that throws an exception in order to unwind the JIT'd 120code's stack. 121 122Now we can construct our CompileOnDemandLayer. Following the pattern from 123previous layers we start by passing a reference to the next layer down in our 124stack -- the OptimizeLayer. Next we need to supply a 'partitioning function': 125when a not-yet-compiled function is called, the CompileOnDemandLayer will call 126this function to ask us what we would like to compile. At a minimum we need to 127compile the function being called (given by the argument to the partitioning 128function), but we could also request that the CompileOnDemandLayer compile other 129functions that are unconditionally called (or highly likely to be called) from 130the function being called. For KaleidoscopeJIT we'll keep it simple and just 131request compilation of the function that was called. Next we pass a reference to 132our CompileCallbackManager. Finally, we need to supply an "indirect stubs 133manager builder": a utility function that constructs IndirectStubManagers, which 134are in turn used to build the stubs for the functions in each module. The 135CompileOnDemandLayer will call the indirect stub manager builder once for each 136call to addModuleSet, and use the resulting indirect stubs manager to create 137stubs for all functions in all modules in the set. If/when the module set is 138removed from the JIT the indirect stubs manager will be deleted, freeing any 139memory allocated to the stubs. We supply this function by using the 140createLocalIndirectStubsManagerBuilder utility. 141 142.. code-block:: c++ 143 144 // ... 145 if (auto Sym = CODLayer.findSymbol(Name, false)) 146 // ... 147 return CODLayer.addModuleSet(std::move(Ms), 148 make_unique<SectionMemoryManager>(), 149 std::move(Resolver)); 150 // ... 151 152 // ... 153 return CODLayer.findSymbol(MangledNameStream.str(), true); 154 // ... 155 156 // ... 157 CODLayer.removeModuleSet(H); 158 // ... 159 160Finally, we need to replace the references to OptimizeLayer in our addModule, 161findSymbol, and removeModule methods. With that, we're up and running. 162 163**To be done:** 164 165** Chapter conclusion.** 166 167Full Code Listing 168================= 169 170Here is the complete code listing for our running example with a CompileOnDemand 171layer added to enable lazy function-at-a-time compilation. To build this example, use: 172 173.. code-block:: bash 174 175 # Compile 176 clang++ -g toy.cpp `llvm-config --cxxflags --ldflags --system-libs --libs core orc native` -O3 -o toy 177 # Run 178 ./toy 179 180Here is the code: 181 182.. literalinclude:: ../../examples/Kaleidoscope/BuildingAJIT/Chapter3/KaleidoscopeJIT.h 183 :language: c++ 184 185`Next: Extreme Laziness -- Using Compile Callbacks to JIT directly from ASTs <BuildingAJIT4.html>`_ 186